This github is the repository for all the assignments for the University of Toronto Course: Computational Design and Fabrication.
Please use the issue tracker to report problems with the assignment code.
Under the hood this code uses the LibFive library for shape manipulation https://libfive.com.
You will notice that methods in these assignment take in variables of the type Kernel::Tree (i.e Kernel::Tree A).
The code treats A as an implicit function A(x). At runtime every x in the scene is passed to A and the surface is evaluated by collecting all x's for which A(x) < 0.
You manipulate shapes by crafting new functions. For instance, let's say you wanted to create a new shape H, that involved the sum (this is different from the union) of two other shapes A and B. You would right
H = A + B
For any given x, the value of H will be A(x) + B(x).
You can see examples of more manipulations on the libfive homepage or in the shapes.cpp in the utils directory.
Libfive supports most standard mathematical operations (+,-,max, min,sin,cos .... ).
For those of you inspired to learn more about optimization by Assignment 2, there are two must read books on the subject (in my opinion). The first is Convex Optimization by Stephen Boyd which is widely considered the best textbook on the subject. The second, and my personal favourite, is Numerical Optimization by Jorge Nocedal and Stephen J. Wright which is more numerically focused and offers a wonderfully clear treatment of common descent based methods for continuous optimization.